AI Politicians: are they desirable?

[Luziyca]

In recent times, some part of me sometimes feels that maybe I should go look for another major besides political science: I may be passionate about that, but what job will that degree bring you? To which I say, "I'm only taking that because when humans need not apply, I don't really want to be one of the lowly jobs."

As for why I feel that I am so confident that artificial intelligence people cannot really be politicians, it is because even if they are more effective than a regular politician, I imagine that many many many people prefer a politician who has actual flesh and blood than a robotic politician, who could be much more effective in its position. As well, if AI began to seriously threaten the politician's positions, they would probably be restricted from entering the political area, or if they are allowed, only to some low-level bureaucratic positions behind-the-scenes.

Therefore, my fellow Generalites, do you believe that artificial intelligence will ever become politicians? Or will they just be rejected because of human demand?

[36 Camera Perspective]

How would I differentiate between one AI politician and another? Why should I support computer A over computer B? Would I have to ask what algorithm each AI uses to calculate its decisions?

[Luziyca]

How would I differentiate between one AI politician and another?

I assume that they will be differentiated by whoever makes the AI.

Why should I support computer A over computer B?

The AI politicians would make promises that would tailor to the ideology of their creators/their own ideology, and what the people want at that exact instant.

Would I have to ask what algorithm each AI uses to calculate its decisions?

No. Do you ask your representative/Senator/legislator how they make their decisions?

[Anime North America]

Honestly it sounds like somethin' out of a bad sci-fi novel. AI is dandy for helping humans (in fact the entire North American economy runs on a decentralized neural network), but AI politicians sounds like a disaster waiting to happen.

[36 Camera Perspective]

I assume that they will be differentiated by whoever makes the AI.

I meant "differentiate in terms of decision-making". That was just a redundant and ambiguous lead-in to my second question.

The AI politicians would make promises that would tailor to the ideology of their creators/their own ideology, and what the people want at that exact instant.

That sounds like it could be inflexible.

No. Do you ask your representative/Senator/legislator how they make their decisions?

Why wouldn't I? That sounds like a perfectly reasonable thing to ask. In any case, I have to wonder, is the goal here to make AI politicians that are superior to human politicians? If so, what would they do better?

[Arkeyana]

If/when we gain sentient AI, they should be given equal rights just like anyone else, so yes.

[Luziyca]

I assume that they will be differentiated by whoever makes the AI.

I meant "differentiate in terms of decision-making". That was just a redundant and ambiguous lead-in to my second question.

Oh. Well, based on the programming of said AI politicians, they could program them to decide differently on the same situation depending on how they programmed the AI politician to function in the political arena, so to both maximize its longevity in office, and to ensure that it governs well.

The AI politicians would make promises that would tailor to the ideology of their creators/their own ideology, and what the people want at that exact instant.

That sounds like it could be inflexible.

It could be inflexible, but if it were to keep tabs on what the people want at that exact millisecond regarding a particular issue, and address it in accordance with its ideology while still pleasing the people, even if they have to betray their "radical ideological principles," then it could be flexible.

No. Do you ask your representative/Senator/legislator how they make their decisions?

Why wouldn't I? That sounds like a perfectly reasonable thing to ask. In any case, I have to wonder, is the goal here to make AI politicians that are superior to human politicians?

That is fair.

But yes, the goal is to make AI politicians who are superior than human politicians, by not only having higher support than human politicians, but also passing legislation that benefits the most people in their jurisdiction.

[Luziyca]

If/when we gain sentient AI, they should be given equal rights just like anyone else, so yes.

They should, but in practice, even if one has equality of opportunity, few are able to take advantage of these opportunities. I feel that AI politicians would, if permitted to run for office, have an ice cube's chance in hell of winning... at least in my lifetime.

[36 Camera Perspective]

-post-

So are we just assuming that the relevant equations/programming exist? For me, the biggest problem with AI politicians is that they would have to operate from some kind of pre-determined equation or algorithm in order to make political decisions, but something like politics is far too dynamic and multicausal to be encompassed effectively in an equation. What kind of equation would describe whether or not to ask Al Franken to step down for sexual harassment accusations, for instance? That does not seem like the kind of thing that can be usefully calculated.

[The Republic of Christiandom]

In recent times, some part of me sometimes feels that maybe I should go look for another major besides political science: I may be passionate about that, but what job will that degree bring you? To which I say, "I'm only taking that because when humans need not apply, I don't really want to be one of the lowly jobs."

As for why I feel that I am so confident that artificial intelligence people cannot really be politicians, it is because even if they are more effective than a regular politician, I imagine that many many many people prefer a politician who has actual flesh and blood than a robotic politician, who could be much more effective in its position. As well, if AI began to seriously threaten the politician's positions, they would probably be restricted from entering the political area, or if they are allowed, only to some low-level bureaucratic positions behind-the-scenes.

Therefore, my fellow Generalites, do you believe that artificial intelligence will ever become politicians? Or will they just be rejected because of human demand?

I, for one, welcome our AI overlords.

[Luziyca]

-post-

So are we just assuming that the relevant equations/programming exist? For me, the biggest problem with AI politicians is that they would have to operate from some kind of pre-determined equation or algorithm in order to make political decisions, but something like politics is far too dynamic and multicausal to be encompassed effectively in an equation. What kind of equation would describe whether or not to ask Al Franken to step down for sexual harassment accusations, for instance? That does not seem like the kind of thing that can be usefully calculated.

Admittedly, explainability via this thread would be somewhat difficult, so alas, I will have to link you to a couple videos about how machines learn: genetic "breeding" model and neural networks.

But considering that AI plays a big role in the stock markets across the world, with trillions of trades a day, they could be used as a starting point for constructing an AI politician.

[The United Remnants of America]

All tests so far that include using a neural network or whatever to basically make decisions as if they were in an elected position result in said program basically becoming a fascist.

So maybe?

[Reutoa]

So are we just assuming that the relevant equations/programming exist? For me, the biggest problem with AI politicians is that they would have to operate from some kind of pre-determined equation or algorithm in order to make political decisions, but something like politics is far too dynamic and multicausal to be encompassed effectively in an equation. What kind of equation would describe whether or not to ask Al Franken to step down for sexual harassment accusations, for instance? That does not seem like the kind of thing that can be usefully calculated.

Admittedly, explainability via this thread would be somewhat difficult, so alas, I will have to link you to a couple videos about how machines learn: genetic "breeding" model and neural networks.

But considering that AI plays a big role in the stock markets across the world, with trillions of trades a day, they could be used as a starting point for constructing an AI politician.

Just like how I will never trust any intelligent AI, I'll never trust or vote for a AI Politician. The real ones are bad enough.

[Luziyca]

Admittedly, explainability via this thread would be somewhat difficult, so alas, I will have to link you to a couple videos about how machines learn: genetic "breeding" model and neural networks.

But considering that AI plays a big role in the stock markets across the world, with trillions of trades a day, they could be used as a starting point for constructing an AI politician.

Just like how I will never trust any intelligent AI, I'll never trust or vote for a AI Politician. The real ones are bad enough.

Even if the AI politicians are superior to real politicians?

[Reutoa]

Just like how I will never trust any intelligent AI, I'll never trust or vote for a AI Politician. The real ones are bad enough.

Even if the AI politicians are superior to real politicians?

Nothing is perfect and everything messes up

[Ifreann]

I, for one, welcome the benevolent rule of Culture Minds.

[Wansul]

AI Technocracy is literally the most diserable future. Future. Not right now. But in the future. It will simply be uncorruptable, efficient, and if coded well, be beneficial to all humans.

[Fartsniffage]

I, for one, welcome the benevolent rule of Culture Minds.

Meatfucker wants a word with you.

[Ifreann]

I, for one, welcome the benevolent rule of Culture Minds.

Meatfucker wants a word with you.

Well it knows my number, I'm sure.

[Infected Mushroom]

Yes.

Why not?

Let's give it a shot.

At least you have some guarantees that their campaign promises will be real (presumably they would be programmed to do that... and if it fails, someone can sue the maker of the AI). Presumably they could be programmed to be 100% "Communist" or "Liberal" and we could set how that works out in a legislative context. There would be no under the table deals where the politician switches sides or wavers so he gets more allies, wins the next election etc... We could program it to want to win elections, but in an ethical and honest way, and then once it gets into power program it to follow its actual platform (as opposed to re-calculating to win the following election).

And we'd all know exactly how the AI would play because we'd all be able to see the programming beforehand (though only some of us would understand it, it can still be layed out to us).

[The Alma Mater]

Even if the AI politicians are superior to real politicians?

Nothing is perfect and everything messes up

Only humans are imperfect. Destroy all humans and the AI can rule over a perfect and logical society.

[Petrolheadia]

Even if the AI politicians are superior to real politicians?

Nothing is perfect and everything messes up

Red fallacy alert!!!

Danger type: perfect solution fallacy.

[Thaioggdol]

No, in the same way, I don't want extraterrestrial leaders ruling Earthly affairs. Maybe 3000 years into the future, but I cannot see me wanting it in the near future.

[Destructive Government Economic System]

What if the AI itself got hacked? If that occurred, then we would be no better off than we already are with modern politics.

[Washington Resistance Army]

Fuck no.